Good day folks.

I am an electronics enthusiast and a full-time musician and mixing engineer.

I had the good fortune of finding a well preserved Teac a3440s reel to reel tape machine.
These machines are considered pro consumer and were built with a pre and post emphasis eq (NAB) to compensate for the flux and valve stages of the day.

To bring this machine up to spec with modern tape manufacture and make it more flexible in a hybrid studio set up I would like to alter that eq.

Currently there is a premphasis at 50hz this happens in the record amp circuit here there is also a cut at 3150hz. On the play amp circuit there is the corresponding 50hz cut and a boost from 3150hz. I would like to rid the circuit of the 50hz Pre and post emphasis and shift the high frequency band to boost and cut at 6300hz in the respective circuits.

Things to note on these circuits would be the high and low speed options which give the operator two tape playback speed options and slight variations on the eq settings. This machine will only run at the high speed setting in my studio.

On the play amp circuit I have noted that the eq shift is altered by a variable resistor and assume that no circuit alterations for the treble fix would be required there. I also note that the treble variations on the record amp with the two speed options is merely a resistor and capacitor difference and assume that a replacement of the “high“ lines R211 and C154 could get me the 6300hz slope desired.

I am handy with a soldering iron and know my way around the machines circuit boards and adjustment controls but circuit design is well out of my depth.

If anyone could give some advise it would be greatly appreciated.

 

It is indeed an RC time constant arrangement and to raise the frequency you can try a lower capacitance and a smaller resistance value. But since there are other RC portions of the circuit the results may not be quite as anticipated. And one note of caution, different capacitor materials have different dielectric absorbtion and leackage properties, and so I suggest staying with the same type as you adjust values.

And it is neat to know that somebody else still considers reel to reel machines worth using. Mine are consumer-grade from the sixties, back when music was good.

 

Thanks for the confirmation and tips Bill, I may start by actually swapping the high and low lines, RC combinations and see where that gets me then I get to keep the exact components in chain, any idea where that 50hz shelf may be implemented?

From further reading it seems all the eq curves have a bass shelf but usually in the 10-20hz range well out of the audio range. Would be great to drop this ones shelf to that range too.

 

As I studied the circuit some more, I see another transistor switch, Q121, with C150 and R209, across the input in the High seed mode.I am not sure just what they do, since they are before the level control.
For the low frequency end, the only component that would reduce the gain as frequency dropped is L102. But since that is in series with the high frequency components the effect may be complicated.
For the playback amplifier some of the EQ circuitry is not shown, below the bottom of the picture.

 

Hi Bill,

Thanks again for your insights, after further digging I have found a couple interesting schematics to look at.

The First two are the Tascam 22-4 which was the first “pro” multitrack recorder from their new teac owners, as we can see it is just the same machine with a little work done. They omitted the portion around Q121 as you suspected it seems unnecessary.
The next two pics are a slightly more modern machine with a switchable eq spec. In the second (record amp) of these something to note is the jumper above ‘u102 2/2’ in its feedback loop, could this be our bass boost?
I am starting to think that the bass cut on the playback side may also have something to do with the fb loop, this is an entirely different circuit but I can’t help notice the jumper being the only difference between the two eq settings. One which has the Nab bass boost and one which doesn’t.

 

The jumpers are indeed affecting the gain in a frequency dependent manner, and indeed thety are in the feedback loop. The design is really quite brilliant in that as the impedance drops with frequency so does the feedback, and so the gain rises. This makes the gain equation more complex, unfortunately I have no idea as to what the IEC equalization curves look like, could they possibly be what you are seeking?

 

Sort of yes, I’d like to run my reel at 15ips with no bass boost as per iec1 standard, but instead of the +-4000hz Lift on iec I’d like a custom 6300hz lift.

 

The current curve On the record side looks like the picture with bass boost and a 3db lift around 3khz. The play side of the circuit would mirror this pre emphasis with cuts.
Iec curve has no bass boost and a 3db Lift around 4.1khz.

 

Currently there is a premphasis at 50hz this happens in the record amp circuit here there is also a cut at 3150hz. On the play amp circuit there is the corresponding 50hz cut and a boost from 3150hz. I would like to rid the circuit of the 50hz Pre and post emphasis and shift the high frequency band to boost and cut at 6300hz in the respective circuits.

Moving the corner frequency to 6.3 kHz will make the system non-interchangeable with anything recorded anywhere else. Why do you want to move it, and where did the 6.3 kHz number come from?

The National Semiconductor Audio Handbook has an excellent section on tape system circuits.

http://bitsavers.trailing-edge.com/components/national/_dataBooks/1976_National_Audio_Handbook.pdf

If you back up through the url you will find most of the National bibles.

ak

 

Unfortunately I cannot give you any advice, but would like to congratulate you for this wonderful machine.
And to record music the way it was before ProTools or Beat Detective.
When correcting a mistake meant cutting and splicing the tape.

 

Years ago I was involved in a fight about using R-R machines the way they were made, verses Dolby verses dbx.
dbx compresses then expands to get more dynamic range. Needs to be recorded and played back on a dbx system.
In my opinion Dolby is the same thing but only works on the high end, where is is needed most. Records in Dolby but sounds "ok" if played back with out Dolby.
I also played with high end expanders that reduce the gain if there is no audio. Really helps the noise and is on the playback end only. (one of the best turned down the treble if there was no high frequencies, which helps the tape hiss)

I was just using my 8-track digital recorder today. It has very good signal to noise ratio. I gave up on my R-Rs when I got a 16 track digital recorder.

 

Hi AK

Yeah this would make it a bit of a custom machine, the reason for the tweaks are purely so I get to use it more, it is not used as a “playback” machine but more as an outboard processor. If I edit 2 of the 4 tracks to the 6300 curve I have two ”hifi” Channels and 2 “lofi“.
The idea of the 63 curve comes up in a J Mcnight paper from MRL tapes although he doesn’t specify this to be a holy grail he merely states that due to the modern formulas and consistency in current tape manufacture that the 6 and 9db+ tapes would offer less noise and more dynamic range if there were no bass boost and a later high boost. Thanks for the link I’m always keen to expand my knowledge.

Thanks for checking in Ron and Schmitt, don’t praise me too much just yet as my studio does run a hybrid setup where I try to utilise the flexibility of the digital and the beautiful soft compression of all the analogue bits. The ultimate goal is to try wrap the audio around the listener while making the process as fun for myself as possible. The more VU and gain reduction LED strips and transformers I can shove in there the better. It’s a low key disco in my mixing room.

I found a second head assembly on eBay and plan on sticking two monsterous DIN two track heads on it so I end up with 3 machines in one.

 

@ronsimpson This is my cassette baby. http://www.hifi-classic.net/review/technics-rs-b100-460.html

dbx does nothing at 1kHz, so that's the test signal you need when putting dbx in service. I had a couple of external dbx decoders for car audio.